Advanced dwarf mongoose optimization for solving CEC 2011 and CEC 2017 benchmark problems

Area of Science:

• Optimization Algorithms
• Computational Intelligence
• Swarm Intelligence

Background:

• The Dwarf Mongoose Optimization (DMO) algorithm faces limitations in convergence rate, particularly when initial solutions are near the global optimum.
• This necessitates modifications to improve its efficiency in complex optimization tasks.

Purpose of the Study:

• To introduce the Advanced Dwarf Mongoose Optimization (ADMO) algorithm, an enhanced version of the DMO.
• To address the convergence rate limitations of the DMO by integrating additional social behaviors of dwarf mongooses.

Main Methods:

• The ADMO algorithm incorporates predation, mound protection, reproductive, and group splitting behaviors observed in dwarf mongooses.
• Modifications were made to the lifestyle of alpha and subordinate groups, as well as foraging and seminomadic behaviors within the DMO framework.
• The performance of ADMO was evaluated using the Congress on Evolutionary Computation (CEC) 2011 and 2017 benchmark functions, including 30 classical/hybrid problems and 22 real-world optimization problems.

Main Results:

• ADMO exhibited improved exploration and exploitation capabilities compared to the original DMO.
• Statistical analysis and performance metrics showed that ADMO achieved superior solutions in most tested benchmark and real-world optimization problems.
• The proposed algorithm outperformed the DMO and seven other established optimization algorithms.

Conclusions:

• The ADMO algorithm effectively overcomes the convergence rate limitations of the DMO.
• The integration of diverse dwarf mongoose social behaviors significantly enhances the algorithm's optimization performance.
• ADMO presents a promising advancement for solving complex optimization challenges in various domains.